EDF Thrust References and Measurement

[telnar1236]

When building EDF jets, a constant problem is figuring out how much thrust I will or won't have. The measured thrusts from manufacturers are frequently very wrong, or measured under conditions that do not represent what we can actually do in an RC plane (e.g. constant voltage at battery pack max charge with no sag). Therefore, I wanted to create a thread where I (and hopefully others) can document what thrust values we actually get out of our EDFs.
These are the ground rules for what I will be doing:

1. All EDF measurements are welcome, regardless of how scientific they are. Even a ballpark number is frequently better than what is now available
2. Measurements must list the EDF unit used at a bare minimum. Ideally listing the battery and ESC will make it easier to judge what is going on
3. Measurements should list anything included beyond the stock EDF (e.g. if it is installed in a plane or has a thrust tube or if tape has been used to reduce the clearance between the blades and duct wall - thanks @L Edge for the suggestion)
4. Measurements will be rated for accuracy as follows
   1. Scientific - made with a purpose-built measurement device, ambient temperature, pressure, current draw, and pack voltage are all listed - I can't achieve this yet, but if someone can, that makes it possible to scale performance and get an idea of how an EDF will work for you in particular
   2. Accurate - made with a reasonable measuring device or better - e.g. an L-shaped lever pushing down on a scale with the EDF on the other arm - and with pack, ESC, and approximate altitude above sea level listed - this should be achievable for most people who want to try and get good measurements
   3. Good - made with a reasonable measuring device or better and EDF listed but the other equipment is not listed
   4. Poor - a rough approximation - e.g. the EDF blowing on a scale and measuring the force of the air - or a rough approximation - e.g. the EDF has enough thrust to fly a plane this heavy

If anyone has EDF thrust values they would like to list here, I would really appreciate any contributions. I'll edit the post following this one to keep an up-to-date list of thrust values across various EDF units as well as add my own measurements in as I make them. I also would guess there might be some variability between individual EDF units, so even if a unit is listed, more measurements make everything more accurate.
Also, if anyone has suggestions for how to improve measurements or thoughts on how to go about this, any thoughts are welcome.

 

[telnar1236]

Thrust Measurements
28 mm EDF
30 mm EDF
40 mm EDF
50 mm EDF
64 mm EDF
70 mm EDF
Powerfun 70 mm 4s EDF with 3400 kv outrunner motor
1800 g thrust - 77 A - accurate - fully charged 4000 mAh 4s HRB lipo and 60 A E Flite ESC - stock with bell mouth installed, no thrust tube
Efficiency: 1.39 g/W

X-Fly 70 mm 6s EDF with 2200 kv inrunner motor
2090 g thrust - 56 A - accurate - fully charged 5000 mAh 6s HRB lipo and 80 A Skywalker ESC - stock (bell mouth is integral to duct), no thrust tube
Efficiency: 1.48 g/W

Freewing 70 mm 6s EDF with 2210 kv inrunner motor
2470 g thrust - 78 A - accurate - fully charged 5000 mAh 6s HRB lipo and 80 A Skywalker ESC - stock with bell mouth installed, no thrust tube
Efficiency: 1.26 g/W

QF Motor 70mm 4s EDF with 2600 kv outrunner motor
1500 g thrust - 59 A - fully charged 4000 mAh 4s HRB lipo and 60 A E Flite ESC - stock with bell mouth installed, no thrust tube
Efficiency: 1.51 g/W

80 mm EDF
FMS 80mm 6s EDF with 2100 kv inrunner motor
3261 g thrust - 103 A - fully charged 4000 mAh 6s HRB lipo and 100 A Predator ESC - stock (bell mouth is integral to duct), no thrust tube
Efficiency: 1.26 g/W

90 mm EDF
120 mm EDF

 

[quorneng]

Are these thrust figures to be for a "bare" EDF? With or without a bell mouth?
Unless you know the details of the actual EDF installation any thrust figure may not be representative of what might be expected.in a plane.

Some basic applications of duct flow theory can make a significant difference to an EDF's performance.
I find it disappointing just how poor EDF duct systems tend to be particularly on the inlet side.

An example of a computer generated inlet duct to ensure a constant duct area with the minimum of disturbance from direction change.

 

[LitterBug]

Thrust figures also change with static vs moving air. Input and exhaust can both influence these numbers. Generally speaking, the output diameter should be reduced by the diameter of the motor in the EDF and be done gradually over a length of at least x times the diameter of the EDF. (can't remember what x is. may be 2 or 2.5)

 

[telnar1236]

Are these thrust figures to be for a "bare" EDF? With or without a bell mouth?
Unless you know the details of the actual EDF installation any thrust figure may not be representative of what might be expected.in a plane.

Some basic applications of duct flow theory can make a significant difference to an EDF's performance.
I find it disappointing just how poor EDF duct systems tend to be particularly on the inlet side.

An example of a computer generated inlet duct to ensure a constant duct area with the minimum of disturbance from direction change.
View attachment 246755

Honestly, any measurements are welcome, but the ones I plan on taking will be for a "bare EDF." In terms of the bell mouth/limitations of the inlet designs on a lot of planes, there isn't much I can do about that apart from what I noted in the 3rd ground rule. If the inlet design/EDF configuration is noted, it should give a better idea of what was actually measured. The tests I'll be doing will be with the EDF out of the plane and with no thrust tube since it's still a good baseline for comparison. If there is a bell mouth it will be installed (maybe I should add photos of the test setup for each EDF?). It's definitely a good point that I should note if there is a bell mouth inlet installed since a lot of EDFs have an optional one. A lot of the time you can approximate beautifully shaped ducting like you're showing with brute force by adding a ton of inlet area. Sometimes that's the only option, but even then you can improve the flow with various bits of geometry. Here's the cheater inlet design on my F-106. In this case, properly sized inlets would have been way too big and the plane itself was as big as it could get and still fit horizontally across my back seat. The vanes give a significant improvement to the thrust I get.

Thrust figures also change with static vs moving air. Input and exhaust can both influence these numbers. Generally speaking, the output diameter should be reduced by the diameter of the motor in the EDF and be done gradually over a length of at least x times the diameter of the EDF. (can't remember what x is. may be 2 or 2.5)

Definitely agree. Using the static thrust isn't so much because I want to as because I have to (no wind tunnel sadly). However, it is still a useful value and provides a baseline that can be compared. I think there are a few different rules of thumb and different EDFs seem to like different exhaust areas. Mine is 85% of the area of a circle with the fan diameter as a starting point or 90% FSA depending on how much math I feel like doing. It seems to give a good balance of static thrust and maintaining thrust at higher speeds. This is the thrust tube on my F-106 - like you were saying, it narrows down a bit towards the end which increases the efflux speed.

With some EDFs, I have also notice that maintaining close to constant area on the thrust tube, similar to what Quorneng was saying for inlets seems to help.

However, since different people might want to tune their EDFs differently, I will be starting out with measurements with no thrust tube. Eventually I do want to experiment with different thrust tube areas since, like you said, it can affect the thrust significantly, but first I want to get a bunch of reliable numbers for a bunch of EDFs in one place.

 

[telnar1236]

Measured the thrust from my 80mm FMS fan today and the amount of power it puts out is ridiculous with 3.26 kg of thrust and pulling almost 2.6 kW. It made the whole table the thrust stand is mounted on flex a bit and was blowing stuff around behind it. I think I'll need to find a way to secure the window shades on the wall across the room since I'm worried it will break them.
I have also added in efficiency numbers for each EDF since we want at least adequate flight times as well as performance. I was expecting the 6s units to do much better, but it looks like it's about the same between 4s and 6s. The standout winners are the XFly 70mm unit and QF motor 70mm unit both of which get efficiencies of about 1.5 g/W.

 

[quorneng]

telnar1236
Your 1.5g/W efficiency is of course the problem with EDFs. At full power and with a practical battery weight the endurance is going to be limited. For "small" EDFs the problem is even worse.
What this does highlight is the importance of airframe weight.
Many full size jets can carry a fuel load equal to the weight of the bare airframe. Given that the battery is the "fuel" in electric how many EDFs can achieve this sort of ratio despite the structural strength to weight advantage of being a smaller size?
Electric will never achieve the performance/endurance ratio of air breathing liquid fuels but we should be able to do better than a few (5?) minutes.

 

[quorneng]

telnar1236
I note you commented

properly sized inlets would have been way too big

I try to overcome this issue by using an EDF that is sized for the inlet and making the plane lighter to compensate for the lower thrust available.
If successful the thrust to weight ratio san remain the same. Modern powerful EDFs can help although their g/W efficiency tends to negate any airframe weight saving.
Now if a 5+ g/W EDF efficiency could be achieved........I wonder.

 

[telnar1236]

telnar1236
Your 1.5g/W efficiency is of course the problem with EDFs. At full power and with a practical battery weight the endurance is going to be limited. For "small" EDFs the problem is even worse.
What this does highlight is the importance of airframe weight.
Many full size jets can carry a fuel load equal to the weight of the bare airframe. Given that the battery is the "fuel" in electric how many EDFs can achieve this sort of ratio despite the structural strength to weight advantage of being a smaller size?
Electric will never achieve the performance/endurance ratio of air breathing liquid fuels but we should be able to do better than a few (5?) minutes.

Definitely agree. And the better specific energy of jet fuel doesn't hurt either - something like 10x better than a LiPo battery. And as the jet burns fuel it gets lighter and needs less power. To your point, a very light efficient airframe can get far better flight times. I recently bought the 64mm L-39 from Banana Hobby and have been very impressed. The battery is about 600g in an airframe with a 630g empty weight and it has efficient straight wings. And it gets 10 minutes of flight time pretty easily without needing to worry that much about throttle management. I'd love to adapt that power system for my own designs.

telnar1236
I note you commented

I try to overcome this issue by using an EDF that is sized for the inlet and making the plane lighter to compensate for the lower thrust available.
If successful the thrust to weight ratio san remain the same. Modern powerful EDFs can help although their g/W efficiency tends to negate any airframe weight saving.
Now if a 5+ g/W EDF efficiency could be achieved........I wonder.

The big issue with this design is that I wanted it to fit in my car's back seat without folding the seats or in the trunk without folding the seats. It puts a very hard upper limit on the size. But going down from a 70mm fan to a 64mm fan would have cut the thrust in half at best. With the limits of 3D printing filaments, I wasn't going to be able to keep it light enough to fly with a smaller fan. Hence the cheater inlets. I would like it if I could keep the weight lower, but with the available filaments, I can only keep planes so light.

 

[quorneng]

It does indeed come down to the materials used.
At the bottom line 3D printing a airframe is a technical achievement but as a process for a complete airframe it is not the lightest option.
Light weight requires the appropriate material for each part depending on the duty it is under. Taken to the extremes it is a slow and laborious process but at model sizes any practical weight saving is limited by the individual part size.
In most cases the 'skin' of a plane that is least suited to be a solid material as it is thin and has a big area. A cell foam is better suited, however there are components where the loads involved do benefit from using a high strength material formed into in a complex shape.
A rather extreme example of using this technique.
Printed fuselage formers with a thin foam skin.

It makes a Stiff and light fuselage. The wings are done in a similar way. Only the EDF nacelles and pylons are fully 3D printed.
Note a plane that uses modern big diameter turbofans has scale inlets plenty big enough for an EDF or in this case even a ducted prop.
You do however have to be rather dedicated to "weight saving" to consider doing it like this!

 

[telnar1236]

It does indeed come down to the materials used.
At the bottom line 3D printing a airframe is a technical achievement but as a process for a complete airframe it is not the lightest option.
Light weight requires the appropriate material for each part depending on the duty it is under. Taken to the extremes it is a slow and laborious process but at model sizes any practical weight saving is limited by the individual part size.
In most cases the 'skin' of a plane that is least suited to be a solid material as it is thin and has a big area. A cell foam is better suited, however there are components where the loads involved do benefit from using a high strength material formed into in a complex shape.
A rather extreme example of using this technique.
Printed fuselage formers with a thin foam skin.
View attachment 247340
It makes a Stiff and light fuselage. The wings are done in a similar way. Only the EDF nacelles and pylons are fully 3D printed.
Note a plane that uses modern big diameter turbofans has scale inlets plenty big enough for an EDF or in this case even a ducted prop.
You do however have to be rather dedicated to "weight saving" to consider doing it like this!

Yeah, unfortunately 3D printing is awful when it comes to keeping the weight down. Or to put it another way printers aren't precise enough to print structures that would be both rigid enough not to buckle and thin enough to take advantage of the strength of the filament. With bigger airframes, the penalty is a lot lower, especially using ABS with its lower density, but with small planes and PLA you end up with flying bricks.

For me, I think the biggest reason to 3D print is that it makes it really easy to reproduce work you already did once a crash happens or if you decide to get rid of a plane and then want to build a new one down the road. The plane in that picture is beautiful and as much a work of art as a piece of engineering. But I can't imagine how long it took to build and how many hours you would need to spend repairing it if it crashed. Where with a 3D printed plane, it is a lot heavier and probably flies a bit worse (or a lot worse if poorly designed), but once you have it designed, if something breaks you just push a button and you have a replacement part the next day. I have crashed my modular F-104 several times now, including a few times where the whole airframe has been destroyed, and I can always have it back up and flying within a week or two of printing and with maybe three or four hours actual hands on work on my part.

There's a lot of work up front, but once everything is designed, 3D printed planes are probably the easiest DIY planes to assemble. Incidentally, it's also one of the reasons I like EDFs. Without a prop on the front the electronics pretty much always survive (unless it's the EDF that the F-104 was designed around and then it causes the crashes).

 

[Mr NCT]

Yeah, unfortunately 3D printing is awful when it comes to keeping the weight down. Or to put it another way printers aren't precise enough to print structures that would be both rigid enough not to buckle and thin enough to take advantage of the strength of the filament. With bigger airframes, the penalty is a lot lower, especially using ABS with its lower density, but with small planes and PLA you end up with flying bricks.

For me, I think the biggest reason to 3D print is that it makes it really easy to reproduce work you already did once a crash happens or if you decide to get rid of a plane and then want to build a new one down the road. The plane in that picture is beautiful and as much a work of art as a piece of engineering. But I can't imagine how long it took to build and how many hours you would need to spend repairing it if it crashed. Where with a 3D printed plane, it is a lot heavier and probably flies a bit worse (or a lot worse if poorly designed), but once you have it designed, if something breaks you just push a button and you have a replacement part the next day. I have crashed my modular F-104 several times now, including a few times where the whole airframe has been destroyed, and I can always have it back up and flying within a week or two of printing and with maybe three or four hours actual hands on work on my part. View attachment 247384 There's a lot of work up front, but once everything is designed, 3D printed planes are probably the easiest DIY planes to assemble. Incidentally, it's also one of the reasons I like EDFs. Without a prop on the front the electronics pretty much always survive (unless it's the EDF that the F-104 was designed around and then it causes the crashes).

Your GeeBee is a beauty! (Jets are okay, too )